Meta’s Refreshed AI Play

Apr 9, 2026 21m

AI Summary Available

Get the full experience! Sign up to access transcripts, personalized summaries, and more features.

Episode Description

Meta releases its first model in the great AI horserace catchup race. OpenAI is expecting to make as much money from ads as the biggest social media companies. Meta doesn’t want you to see ads encouraging you to join lawsuits against them. And Amazon’s big Starlink competitor is coming as soon as this summer. Meta releases first AI model since Zuckerberg’s spending spree (FT) AI Horserace Thread (@emollick) Anthropic Completes Tender Offer, But Employees Hold Onto Shares (Bloomberg) OpenAI Forecasts Advertising to Hit $102 billion by 2030 (The Information) Scoop: Meta removes ads for social media addiction litigation (Axios) Amazon’s Starlink competitor Leo gets a new date (The Verge) Deere settles US right-to-repair lawsuit with $99 million fund, repair commitments (Reuters) Gen Z Is Using A.I., but Doesn’t Feel Great About It (NYTimes) Learn more about your ad choices. Visit megaphone.fm/adchoices

AI-Generated Summary

Meta’s New AI Model and Strategy

Meta has launched Muse Spark, its first new AI model from its superintelligence labs, aiming to enhance its products with smarter, faster, and more visual AI responses. The model is trained in health applications with the help of over a thousand doctors and features capabilities like personalized shopping and multimodal perception. Mark Zuckerberg emphasized ongoing investments to develop advanced AI products that act as personal agents, hinting at future open-source models and integrations with wearable devices like AR glasses. Despite being a smaller, closed model, Muse Spark outperformed some top rivals on certain benchmarks, indicating Meta’s renewed focus on niche AI applications amidst the ongoing AI race.

The Global AI Race and Market Leaders

The podcast discusses the current landscape of AI model development, highlighting that US companies like Google, OpenAI, and Anthropic dominate the frontier with closed-source models, some of which possibly self-improve recursively. Meta is re-entering the race with a less advanced but promising approach. Meanwhile, several Chinese companies such as Alibaba and Xiaomi are maintaining momentum, though their models lag months behind US leaders. Questions remain about the progress of self-improvement, reliance on distillation techniques, and the future of open versus closed-weight models, with US labs moving away from open models.

Anthropic’s Valuation and Business Growth

Anthropic, another key player, has recently conducted a secondary share sale, raising concerns as it fell short of its fundraising target. Despite selling some employee-held shares, the company remains highly valued—over $350 billion—thanks to rapid revenue growth, surpassing $19 billion annually. The company’s investors remain optimistic about its prospects, especially as Anthropic prepares for an IPO possibly later this year.

OpenAI’s Advertising Ambitions and Revenue Projections

OpenAI is aggressively expanding its revenue streams through advertising, aiming to generate $2.4 billion this year, rising to nearly $11 billion next year, and potentially $102 billion by 2030—about half of the revenue Meta earns from advertising. The company’s focus on integrating ads into ChatGPT’s ecosystem, with rising ARPU from nearly $3.50 this year to $60 in 2030, reflects its commitment to monetizing the large user base. Despite this, OpenAI continues to grapple with its primary challenge: maintaining user engagement and trust while expanding monetization.

Meta’s Crackdown on Social Media Addiction Ads

Meta has begun removing numerous ads from Facebook and Instagram promoting legal services targeting minors, especially amid ongoing litigation around social media addiction and its effects on youth. The platform relies on its terms of service to restrict certain content and has deactivated many harmful advertisements, although some still remain. This move indicates Meta’s efforts to mitigate legal risks and address public concerns about social media’s impact on young users.

Amazon’s Starlink Competitor, Leo, Launch and Development

Amazon announced that its space-based internet service, Leo (formerly Project Kuiper), is scheduled to launch in mid-2026 after delays. With FCC approval for over 3,200 satellites, Amazon plans to use rideshare launches until its own rockets are ready. Leo promises faster, cheaper internet and seamless integration with AWS for enterprise and government use. The project aims to fill global connectivity gaps, providing an alternative to SpaceX’s Starlink, despite currently having a smaller satellite constellation.

US Agriculture Equipment and Right-to-Repair Lawsuit Settlement

Deere has agreed to a $99 million settlement fund for farmers concerning repair costs and access to diagnostic tools. This settlement resolves aspects of a broader legal conflict over right-to-repair issues, where Deere was accused of restricting farmers’ access to necessary repair tools, leading to higher costs. The deal emphasizes the push for more open repair practices in the agricultural sector, although Deere denies wrongdoing.

Young People’s Attitudes Toward AI

A recent Gallup survey reveals that young people aged 14-29 are increasingly skeptical and less hopeful about AI, with positive attitudes dropping from 27% to 18% over the past year. While many use AI regularly, concerns persist about its impact on creativity, critical thinking, and employment. Especially in the workforce, nearly half believe AI’s risks outweigh its benefits, reflecting a broader resistance to AI’s rapid integration into daily life among the younger generation.